PROJECT SUMMARY Successful targets for cancer therapy fall in three main categories: oncogenes that elicit tumor-specific essentiality because of their direct role in tumorigenesis (oncogene dependency), proteins that elicit synthetic lethality with specific mutations despite lack of a direct role in tumorigenesis (non-oncogene dependency), and proteins related to the interaction of tumor cells with the immune system (immune-checkpoint dependency). However, given our current understanding of cancer as a complex and highly heterogeneous system, it is difficult to imagine that an individual protein may represent an effective target for all the billions of cells that make up a typical mass. Indeed, while genetic-based targeted therapy and immunoncology hold great promise a majority of patients still does not respond or will eventually relapse with drug resistant tumors, suggesting that the concept of therapeutic targets as single proteins may need to be revisited. To accomplish this goal, we will leverage a highly successful framework developed by our center investigators for the identification and pharmacological targeting of tumor dependencies implemented by the concerted activity of a handful of Master Regulator (MR) proteins within tightly regulated tumor checkpoint modules. Specifically, we will elucidate and experimentally validate MR proteins and associated Tumor Checkpoint modules of rare and incurable malignancies, on an individual patient basis, by performing network-based analysis of tumor samples signatures using regulatory models reverse engineered from primary tumor samples. We will then prioritize a set of FDA approved drugs and late stage investigational drugs in phase II or phase III studies in oncology (oncology drugs) based on their ability to either target essential/synthetic-lethal MRs (OncoTarget) or to reverse the full MR signature of a tumor (OncoTreat). RNASeq profiles of appropriately matched tumor models perturbed with available oncology drugs will be obtained and analyzed to assess the differential tumor checkpoint activity induced by individual drugs and drug combinations, followed by low-throughput studies to elucidate the regulatory basis of their activity. Models ? including cell lines, short-term organotypic culture from tumor explants (EXPL), organoids (ORG), and patient derived xenografts (PDX) ? will be selected based on MR protein conservation. We will validate these findings as well as overall efficacy of prioritized drugs and combinations, pharmacodynamic properties, biomarker accuracy and sensitivity, and mechanisms of resistance in suitable in vitro and in vivo models. If successful, this would represent the first mechanistic approach for precision cancer medicine, where therapeutic targets, associated inhibitors, and population stratification biomarkers are systematically derived from precise, mechanistic understanding of tumor state regulation and of its drug-induced modulation.